Search for O{}_{2} Toward the Orion Bar


We report the results of a search for molecular oxygen (O) toward the Orion Bar, a prominent photodissociation region at the southern edge of the HII region created by the luminous Trapezium stars. We observed the spectral region around the frequency of the O 31 transition at 487 GHz and the 53 transition at 774 GHz using the Heterodyne Instrument for the Far Infrared on the Herschel Space Observatory. Neither line was detected, but the 3 upper limits established here translate to a total line-of-sight O column density 1.510 cm for an emitting region whose temperature is between 30K and 250K, or 110 cm if the O emitting region is primarily at a temperature of 100K. Because the Orion Bar is oriented nearly edge-on relative to our line of sight, the observed column density is enhanced by a factor estimated to be between 4 and 20 relative to the face-on value. Our upper limits imply that the face-on O column density is less than 410 cm, a value that is below, and possibly well below, model predictions for gas with a density of 1010 cm exposed to a far ultraviolet flux 10 times the local value, conditions inferred from previous observations of the Orion Bar. The discrepancy might be resolved if: (1) the adsorption energy of O atoms to ice is greater than 800K; (2) the total face-on A of the Bar is less than required for O to reach peak abundance; (3) the O emission arises within dense clumps with a small beam filling factor; or, (4) the face-on depth into the Bar where O reaches its peak abundance, which is density dependent, corresponds to a sky position different from that sampled by our Herschel beams.

astrochemistry – ISM: abundances – ISM: individual objects (Orion) – ISM: molecules – submillimeter: ISM

Herschel Search for O Toward the Orion Bar

*[9mm] Gary J. Melnick, Volker Tolls, Paul F. Goldsmith, Michael J. Kaufman,

David J. Hollenbach, John H. Black, Pierre Encrenaz, Edith Falgarone, Maryvonne Gerin, Åke Hjalmarson, Di Li, Dariusz C. Lis, René Liseau, David A. Neufeld, Laurent Pagani,

Ronald L. Snell, Floris van der Tak, and Ewine F. van Dishoeck

Received ;     Accepted 

 Herschel is an ESA space observatory with science instruments provided by European-led 0.Principal Investigator consortia and with important participation from NASA

  1. Harvard-Smithsonian Center for Astrophysics, 60 Garden Street, MS 66, Cambridge, MA 02138, USA

  2. Jet Propulsion Laboratory, California Institute of Technology, 4800 Oak Grove Drive, Pasadena, CA 91109, USA

  3. Department of Physics and Astronomy, San Jośe State University, San Jose, CA 95192, USA

  4. SETI Institute, Mountain View, CA 94043, USA

  5. Department of Earth & Space Sciences, Chalmers University of Technology, Onsala Space Observatory, SE-439 92 Onsala, Sweden

  6. LERMA & UMR8112 du CNRS, Observatoire de Paris, 61 Av. de l’Observatoire, 75014 Paris, France

  7. LRA/LERMA, CNRS, UMR8112, Observatoire de Paris & École Normale Supérieure, 24 rue Lhomond, 75231 Paris Cedex 05, France

  8. National Astronomical Observatories, Chinese Academy of Sciences, A20 Datun Road, Chaoyang District, Beijing 100012, China

  9. California Institute of Technology, Cahill Center for Astronomy and Astrophysics 301-17, Pasadena, CA 91125, USA

  10. Department of Physics and Astronomy, Johns Hopkins University, 3400 North Charles Street, Baltimore, MD 21218, USA

  11. Department of Astronomy, University of Massachusetts, Amherst, MA 01003, USA

  12. SRON Netherlands Institute for Space Research, P.O. Box 800, 9700 AV, and Kapteyn Astronomical Institute, University of Groningen, Groningen, The Netherlands

  13. Leiden Observatory, Leiden University, P.O. Box 9513, 2300 RA, Leiden, The Netherlands

  14. Max-Planck-Institut f¬ur Extraterrestrische Physik, Giessenbachstrasse 1, 85748, Garching, Germany

1 Introduction

Searches for interstellar O have a long history, but their motivation has evolved with time. Prior to the late-1990’s, efforts to detect O were driven largely by a desire to confirm its predicted role as a major reservoir of elemental oxygen within dense molecular clouds and as the most important gas coolant – after CO – of cold (30K), modestly dense ((H)1010 cm) gas (cf. Goldsmith & Langer, 1978; Neufeld, Lepp, & Melnick, 1995). The launch of the Submillimeter Wave Astronomy Satellite (SWAS) in 1998 and Odin in 2001, and the subsequent failure of these observatories to detect O toward a large number of sources at levels of a few percent of the abundances predicted by equilibrium gas-phase chemical models, have forced a shift in emphasis to a re-examination of the oxygen chemistry in dense molecular gas. Today, interest in O no longer lies in its being a significant reservoir of elemental oxygen or in its cooling power. Instead, because the abundance of gas-phase O is set by a balance of various formation, destruction, and depletion processes thought to affect the broader chemistry in dense gas – such as gas-phase reactions, grain-surface reactions, thermal sublimation, far-ultraviolet (FUV) photodesorption, cosmic-ray desorption, photodissociation, and freeze out – measures of O have become an important test of our current understanding of the relative effectiveness of these processes.

The capabilities of the Herschel Space Observatory’s Heterodyne Instrument for the Far-Infrared (HIFI; de Graauw etal., 2010) have enabled improved searches for O through: (1) its high sensitivity, including at 487 GHz – the frequency of the 31 transition observed previously by SWAS and Odin; and, (2) its broad frequency coverage that permits observations of additional O submillimeter transitions, some of which are expected to exhibit stronger emission than the 31 line under certain physical conditions. The Open Time Key Program “Herschel Oxygen Project” (HOP; Co-PI’s P. Goldsmith and R. Liseau) is designed to survey Galactic sources with the goal to detect O or set meaningful limits on its abundance within these regions. Because the effectiveness of the processes that determine the O column density depends upon the gas density, temperature, and incident FUV flux  (scaling factor in multiples of the average Habing local interstellar radiation field; Habing 1968) among other parameters, testing these models requires that the HOP observations include a range of source types, such as dense quiescent clouds, outflows and shocked gas regions, and FUV-illuminated cloud surfaces (see, for example, Goldsmith etal., 2011; Liseau etal., 2012).

In this paper, we report the results of a deep search for O emission toward the Orion Bar, a well known ionization front located approximately 2 southeast of the Trapezium stars in Orion at the interface of the HII region created by these stars and the dense gas associated with the surrounding Orion molecular cloud. The Orion Bar lends itself well to the study of FUV-illuminated molecular gas for several reasons, including its nearly edge-on geometry, its proximity (420 pc; Menten etal., 2007; Hirota etal., 2007; Kim etal., 2008), its relatively high density ((H)310 cm), and the strong (1010) external FUV field irradiating this gas. The Orion Bar, and sources like it, are of particular interest since the dust grains within these regions are predicted to be sufficiently warm that the thermal evaporation of O atoms from the grain surfaces is enhanced, resulting in a higher fraction of O in the gas phase and the increased production of O via gas-phase chemical reactions (OOHOH). Under such circumstances, the O column density can be more than a factor of 10 greater than within gas exposed to lower (i.e., 500) external FUV fields (cf. Hollenbach etal., 2009). The inclusion of the Orion Bar within the HOP program was intended to test this prediction.

The observations and data reduction methods are described in §2 below. In §3, we present the resultant spectra and the upper limits to the O integrated intensity. In §4, we review the excitation conditions within the Orion Bar and the derived limits on the line-of-sight O column density. In §5, we discuss these limits in the context of recent chemical models that trace the O abundance from the FUV-illuminated cloud surface to the deep interior.

2 Observations and Data Reduction

The Herschel HIFI observations presented here were carried out using the HIFI Band 1a receiver for the 31 487 GHz observations and the HIFI Band 2b receiver for the 53 774 GHz observations. The 487 GHz observations were conducted on operational day (OD) 291 in spectral scan dual beam switch (DBS) mode, while the 774 GHz observations were conducted on OD 297 in spectral scan DBS mode and on OD 509 in HIFI single point DBS mode. Eight LO-settings were used for both the 487 GHz and the 774 GHz spectral scans to enable the spectral deconvolution, and the additional eight single point 774 GHz observations were observed also using eight different LO settings. The total integration time (on-sourceoff-source) for each polarization was 0.93 hours for the 487 GHz spectral scan, 0.86 hours for the 774 GHz spectral scan, and a total of 4.6 hours for the eight single point 774 GHz observations. The full-width-at-half-maximum (FWHM) beam sizes were 44.7 at 487 GHz and 28.2 at 774 GHz.

The observed position, 5h 35m 20.6s, 52514.0Õ (J2000), is shown in Fig. 1. We applied the total observing time allotted to HOP observations of the Orion Bar to a single spatial position – versus multiple positions – in order to achieve the lowest radiometric noise and, thus, the greatest sensitivity to weak O emission. In the absence of prior information about the possible O spatial distribution, our choice of sky position was guided by the desire to place the 487 GHz and 774 GHz beam centers a distance corresponding to approximately 8 visual magnitudes into the molecular gas measured from the ionization front, in accord with model predictions (see §5 for a full discussion). For an H density between 510 cm and 510 cm, applicable to the interclump medium in the Bar, and 10, this corresponds to a projected angular distance of between 2.4 and 24 from the ionization front. As shown in Fig. 1, the selected position places the beams in the center of this range, while the beam sizes encompass the full range. The sky position parallel to the Orion Bar was selected to coincide with the molecular gas, as delineated by the CO 32 emission (see Fig. 1), and, for future analysis, one of the positions under present study by another Herschel Key Program.

The data were processed using the standard HIFI pipeline software HIPE version 7.3 (Ott, 2010), spurious signals (“spurs”) removed, spectra defringed, spectral scans deconvolved, and all data finally exported to GILDAS-CLASS format. Further processing was performed only on the Wide Band Spectrometer (WBS) spectra (0.5 MHz channel spacing and 1.1 MHz effective spectral resolution) using the IRAM GILDAS software package (, including first-order baseline removal, averaging of the 774 GHz spectral scans and frequency-aligned single point observations, averaging of the H- and V-polarization spectra, and production of separate averages for both frequencies and both sidebands. The frequencies for the line identification were extracted from the JPL and CDMS databases (Pickett etal., 1998; Müller etal., 2005) as well as Drouin etal. (2010) in the case of O.

3 Results

A summary of the identified lines in the HIFI Band 1a and Band 2b spectra along with the observing modes, integration times, and Gaussian fit parameters is provided in Table 1. The summed H+V polarization spectra observed in Band 1a are shown in Fig. 2, while those observed in Band 2b are shown in Fig. 3. With the exception of the HCl chloronium 485 GHz spectrum, which is a blend of three hyperfine components (cf. Lis etal., 2010; Neufeld etal., 2011), all of the detected lines appear well fit by single Gaussian profiles with a common LSR line center of 10.680.14 km s (1) and individual best-fit FWHM line widths ranging from about 1.8 km s to 2.5 km s.

The upper limit to the integrated intensity of the O 31 and 53 transitions is derived assuming each line is described by a single Gaussian profile, as is the case for the other unblended lines we detect toward this position. The rms noise in the O 31 487 GHz spectrum between LSR velocities of 110 km s and 25 km s – a velocity range within which there is no evidence for any spectral features – is 2.62mK per 0.5 MHz channel. Similarly, the rms noise in the O 53 774 GHz spectrum between LSR velocities of 70 km s and 30 km s is 2.19mK per 0.5 MHz channel. The intrinsic O line widths along this line of sight are unknown; however, we assume they lie between the extremes of 1.8 km s and 2.5 km s (FWHM) measured for the other unblended lines we detect along this line of sight (see Table 1). This leads to 3 upper limits of between 0.0150 and 0.0209 Kkm s for the 31 487 GHz line and between 0.0126 and 0.0175 Kkm s for the 53 774 GHz line.

4 Excitation and Limits on the O Column Density

The Orion Bar, like many other photodissociation regions (PDRs), displays emission from a variety of ionic, atomic, and molecular species best fit by a mix of gas densities and temperatures. The broad picture to emerge is that of a layer consisting of at least two components: interclump gas with (H32010cm (Hogerheijde, Jansen, & Van Dishoeck, 1995; Wyrowski etal., 1997; Simon etal., 1997; Marconi etal., 1998) surrounding clumps with (H1010cm (Lis & Schilke, 2003; Young Owl etal., 2000), which comprise about 10% of the mass (Jansen etal., 1995). Gas temperature estimates similarly vary, depending on the species observed and the component giving rise to most of the emission. Within the denser well-shielded gas, the gas temperature is thought to range between 50 and 85K (Hogerheijde, Jansen, & Van Dishoeck, 1995; Gorti & Hollenbach, 2002). The gas temperature associated with the interclump medium is estimated to be 8530K (Hogerheijde, Jansen, & Van Dishoeck, 1995), with some gas temperatures associated with the surfaces (A1) of the denser clumps ranging as high as 220K (Jansen etal., 1995; Batrla & Wilson, 2003; Goicoechea etal., 2011). There is evidence for an even warmer component (300700K) based on emission from pure rotational lines of H and far-infrared fine-structure lines of [OI] at 63 and 145m and [CII] at 158(Herrmann etal., 1997; Allers etal., 2005). This warmer component is believed to arise in the gas between the ionization front and the molecular region traced by CO emission (Walmsley etal., 2000). The strength of the FUV field incident on the Orion Bar has been estimated to be 1410 based upon the total radiation from the Trapezium stars – and the O star Ori C in particular – the intensity of the far-infrared [CII] and [OI] fine-structure lines mapped toward the Orion molecular ridge, the strength of several near-infrared lines whose intensities have been ascribed to recombinations to highly excited states of CI, and the strength of near-infrared NI lines excited by the fluorescence of UV lines (Herrmann etal., 1997; Marconi etal., 1998; Walmsley etal., 2000). Given a density of 10 cm for the bulk of the material and a  of 10, models predict that the O abundance peaks at A8 mag. (cf. Sternberg & Dalgarno, 1995; Hollenbach etal., 2009). At these depths into the cloud, the gas temperature is predicted to be 3040K (Hollenbach etal., 2009). Thus, in our analysis, we consider the possibility that the O emission could arise in gas with temperatures anywhere between 30K and 250K.

The weak line flux of the O magnetic dipole transitions makes them highly likely to be optically thin. Under the assumption that the O emission uniformly fills the HIFI beam, the observed integrated intensity in a given transition is:


where is the main beam temperature, is the line frequency (and is the line frequency in GHz), is the spontaneous decay rate between the transition upper level, , and lower level, , (O) is the total O column density in cm, and is the fractional population in the transition upper level. The conversion between main beam and antenna temperature makes use of the efficiencies reported in Roelfsema etal. (2012).

To determine the fractional population of the transition upper state, , the excitation of the lowest 36 levels of O, corresponding to a maximum upper-level temperature of 1141 K, was computed under the large velocity gradient (LVG) approximation. The spontaneous decay rates are those of Drouin etal. (2010) and the collisional rate coefficients are those calculated by Lique (2010) for HeO collisions, multiplied by 1.37 to account for the different reduced mass when H is the collision partner. For molecular hydrogen densities 310 cm, both the 31 and 53 transitions are close to (or in) LTE and the values of depend essentially only on the temperature. Fig. 4 shows the resulting contours of integrated antenna temperature for the 31 transition as functions of the total O column density and gas temperature between 30 and 250 K. Similarly, Fig. 5 shows the corresponding results for the 53 transition.

Of the two O lines searched for here, an examination of Figs. 4 and 5 shows that our measured upper limits to the 53 774 GHz integrated intensity place a more stringent limit on the maximum O column density for 35K (and comparable limits to that set by the 487 GHz line at 30K). Specifically, assuming the emission fills the beam, the total line-of-sight O column density must be less than 1.510 cm (3). If the O abundance peaks within the cooler well-shielded gas, for which 100K, the upper limit to the total O column density is less than 110 cm (3).

5 Discussion

O is produced primarily through the gas-phase reaction OOHOH and is destroyed by photodissociation for the cloud depths of interest here. Thus, the O abundance is expected to peak where the FUV field has been heavily attenuated and where both the gas-phase O and OH abundances are high which, in externally FUV-illuminated clouds, is predicted to occur within a relatively narrow (i.e., a few A deep) zone centered at an A  9 mag. from the cloud surface (cf. Hollenbach etal., 2009). The proximity of this zone to the surface and the range of depths over which the peak abundance occurs are governed by several important processes. Near the cloud surface, where the FUV field is largely unattenuated, the equilibrium O abundance is low owing to a high photodissociation rate. Beyond a few A into the cloud, the FUV field is attenuated, the photodissociation rate reduced, and a region of peak O (and HO) abundance is attained.

Within most clouds with 500, the path to O formation is believed to start with the formation of water ice, HO, on grains, which occurs when O atoms strike and stick to grains long enough to combine with an accreted H atom to form OH and then HO. Within this region the FUV field remains strong enough to photodesorb HO from the ice mantles and subsequently photodissociate these molecules, creating sufficient gas-phase O and OH to produce O by the gas-phase chemical reaction above. Deeper into the cloud (i.e., greater A), the FUV field is almost completely attenuated and the gas-phase OH and HO produced through the photodesorption and photodissociation of HO drops significantly; most O atoms that then strike dust grains and form HO remain locked in ice as long as the grain temperature is 100K. Over time (10 years), this process greatly reduces the gas-phase atomic oxygen abundance and suppresses the formation and abundance of O. Hence, in the model of Hollenbach etal. (2009), the steady-state abundance profile of O (and HO) resembles an elevated plateau that peaks at an A6 for gas with (H)1010 cm and 500.

For regions subject to a  greater than 500, such as the Orion Bar, the scenario above is altered and, for several reasons, the peak O abundance is higher and occurs at a higher A. First, the high FUV field absorbed at the cloud surface leads to a high infrared field that keeps the grains warm, even deep within the cloud. For 10, 40K to A8, resulting in a significant fraction of the O atoms being thermally desorbed from the grains before they can form HO and leading to an increase in O in the gas phase. Second, the higher grain temperature also reduces the freezeout of such oxygen-bearing species as OH and O, further increasing the amount of elemental O in the gas phase. Finally, the attenuated FUV flux at the higher values of A lowers the photodestruction rates, allowing O to survive to greater cloud depths. The combined result of these effects is a peak O abundance about 3 times higher, and a total O column density more than 10 times greater than for comparably dense gas exposed to 500. This result is reflected in the detailed calculations presented in Hollenbach etal. (2009) and shown in Fig. 6, which is adapted from their paper. For this reason, the Orion Bar was considered a promising source for our attempts to detect O emission.

From Fig. 6, it would appear that the upper limits on the total O column density established here are not in serious disagreement with the model predictions. However, the results shown in Fig. 6 apply to a gas column perpendicular to the face of a planar cloud. This is not the geometry of the Orion Bar, which has often been described as an edge-on PDR, though its true structure has been the subject of some study and debate. For example, based on millimeter and submillimeter line observations, Hogerheijde, Jansen, & Van Dishoeck (1995) and Jansen etal. (1995) propose a model in which the Bar has a tilt angle, , of 3 from edge-on, resulting in an increase in the line-of-sight column density (beyond what would be measured for a face-on geometry) by a factor (sin), or almost 20. Alternately, Walmsley etal. (2000) find that a cylindrical model, in which the axis is in the plane of sky and the radius is 0.3pc, best reproduces the observed spatial distribution of the fluorescent OI 1.317m emission. In this scenario, the average geometrical enhancement of the line-of-sight depth into the Bar versus the face-on depth is about 5. Finally, Neufeld etal. (2006) find that a geometrical enhancement factor of 4 is required to reconcile observed and predicted C column densities.

The 3 upper limit to the face-on O column density can thus be inferred from our line-of-sight values to be 1.510sin cm, or 1.010sin cm for 100K. (We note that these upper limits are derived assuming the intrinsic O FWHM line width is 2.5  km s; if the intrinsic width is closer to the lower end of the observed range, i.e., 1.8 km s, the face-on O column density upper limits are further reduced by a factor of 1.4.) For gas densities 10 cm, which applies to most of the gas in the Bar, this is to be compared with a total predicted face-on O column density of 710 cm, as shown in Fig. 6, with most of this column occurring inside a layer of peak O abundance with a width corresponding to approximately 2 magnitudes (see Fig. 7), or a linear size of 1.910/ cm, where (H)/[10 cm]. Viewed from a distance of 420 pc, this zone of peak O emission would subtend 3[(1/ + 162.4sin], where is the physical length of the Bar in parsecs. For 0.6 pc (cf. Jansen etal., 1995) and 1, 6 would result in O emission that fills the Herschel/HIFI beam at 774 GHz, though a minimum geometric enhancement factor of 4, derived from other observations, suggests that does not exceed 15. However, these tilt angles imply an upper limit to the face-on O column density between 1.610 cm and 3.910 cm, which is below, and in some cases, significantly below that predicted by theory.

For 0.6 pc and 1, but 6, the O layer no longer fills the 774 GHz beam. Although the peak O column density within the beam will continue to increase for angles less than 6, the beam filling factor will decrease. These two effects offset exactly, and the beam-averaged O column density will remain the same for all tilt angles less than about 6. Since the O emission is optically thin, the line emission will likewise remain constant within the under-filled beam. In this case, the geometrical enhancement factor would be 10, and the upper limit to the face-on O column density remains below that predicted. Therefore, we conclude that Bar geometry cannot account for the discrepancy between theory and observations.

What, then, can account for the discrepancy? The amount of O produced in externally FUV-illuminated dense gas depends on several factors, which we examine below:

Thermal evaporation: As noted earlier, the dwell time of an O atom on a grain surface can have a considerable effect on the O abundance, particularly when this time becomes less than the time to combine with an H atom on the surface. The timescale for thermal evaporation of an O atom is approximately 910 exp[800K / ] seconds, where 800K is the adsorption energy of O to water-ice (Hasegawa & Herbst, 1993) that applies to van der Waals binding to a chemically saturated surface. It is possible that the binding energy is greater than 800K, which would increase the grain temperature, and thus the , required to thermally desorb O atoms on short timescales and produce the jump in the total O column density for 500 seen in Fig. 6. If, for example, the O adsorption energy was 1600K, grains as warm as 42K – the expected dust temperature at high A in a 10 field – would, on average, retain their O atoms long enough to form HO, thus delaying the 500 rise in O column density seen in Fig. 6 until 10.

Photodesorption yield of HO from a grain surface, Y: The abundance (and column density) of O depends on the gas-phase abundance of O and OH, the latter being produced primarily through the photodissociation of HO, much of which is either photodesorbed from grains or produced via the dissociative recombination of gas-phase HO. At high  (and 20K), short O-atom dwell times on grains suppress the formation of OH and HO. However, even though it is not formed on the grain surface in a high- environment, HO formed in the gas phase via HO dissociative recombination will be depleted through freezeout onto grains and will remain locked in HO for as long as 100K. Since the quantity of OH and HO returned to the gas phase as a consequence of HO photodesorption scales with , the total O column density likewise scales with , as is seen in Fig. 6. A value for  less than 10 would help to reconcile theory and observation. However, fits to the SWAS and Odin HO data (Hollenbach etal., 2009) as well as theoretical simulations and laboratory measurements (Andersson & van Dishoeck, 2008; Arasa etal., 2011; Westley etal., 1995a, b; Öberg etal., 2009) suggest, if anything, that the appropriate value of  is greater than 10.

Grain cross-sectional area (per H): The equilibrium O abundance in the A range of maximum O abundance scales as (), where  is the grain cross-sectional area per H nucleus. Therefore, lowering  will decrease the O column density, bringing model and observation into closer agreement. For an “MRN” (Mathis etal., 1977) grain size distribution , where is the grain radius,  210 cm for an assumed gas-to-dust mass ratio of 100 with grains ranging in radii between a minimum, , of 20Å and a maximum, , of 2500Å (the standard value in Hollenbach etal. 2009). Grains with 20Å will be cleared of ice mantles by single photon heating or cosmic rays and, thus, are not significant ice reservoirs. Because (, in order to lower the value of  while preserving the total mass in grains, either or both  and  must increase, such as through coagulation. For example, a reduction in , and thus the face-on O column density, by at least a factor of 2 could be achieved if the minimum grain radius were to increase to 80Å.

Alternately, the buildup of an ice mantle, which can increase the radius of grains by as much as 50Å, will increase the value of . For values of  of 10 applicable to the Orion Bar, grain temperatures are expected to be 40K, which is high enough to inhibit ice formation via surface reactions (absent a higher O adsorption energy); however, water formed in the gas phase via the reaction HOHOH can still freeze out and form an ice mantle. Toward Orion, there is evidence for a departure from the assumed gas-to-dust mass ratio of 100, which is consistent with the buildup of ice mantles (see, for example, Goldsmith, Bergin, & Lis, 1997). In addition, there is evidence for a deficiency in small grains and for grain growth, possibly due to radiation pressure, the preferential evaporation of small grains, and coagulation (e.g., Cesarsky etal., 2000; Pellegrini etal., 2009; Shaw etal., 2009). The net effect of lowering  through these processes, and increasing  through the accumulation of an ice mantle, is unclear in a high- environment like the Orion Bar.

Beam position: For an interclump H density between 510 cm and 510 cm and 10, the peak O abundance is predicted to occur at a face-on depth into the cloud corresponding to an A8 (see Fig. 7). Thus, the linear distance from the A0 surface, which we assume is the prominent ionization front, to the depth of peak O abundance is 7.610/(H) cm. For an assumed distance of 420 pc, the angular separation between the ionization front and the position of peak O abundance (and column density) is then 1.5A/[(H)/10] arcseconds, where A is the face-on depth of the O peak abundance in magnitudes. Thus, an interclump H density of 10 cm should produce O emission that peaks 12 from the ionization front and close to the center of the observed sky positions (see Fig. 1). However, if the interclump density is more than a factor of 2 different from 10 cm – values that remain within the range of density estimates for the interclump medium – then the peak O abundance is predicted to fall to either side of the observed beam center position.

Finally, we note that the inferred peak line-of-sight H column density, (H), applicable to the interclump medium toward the Orion Bar is estimated to be 6.510 cm (Hogerheijde, Jansen, & Van Dishoeck, 1995). If the geometrical enhancement factor is 10, as would be the case for a tilt angle 5.5, this would imply a face-on H column density of 6.510 cm, corresponding to a total A through the Bar of about 7. If the face-on extinction through the Orion Bar is indeed this low, then the attenuation of the 10 field is not sufficient to allow O to reach its peak abundance and the total O column density will be less than predicted by Hollenbach etal. (2009), whose total column densities are based upon cloud depths corresponding to A10. This is illustrated in Fig. 7, which shows both the profile of O abundance versus A and the cumulative O column density to a given A, computed using the model described in Hollenbach etal. (2009) for the conditions appropriate to the Bar interclump medium. At a depth corresponding to an A of 7, the predicted face-on O column density remains 310 cm, well below the limits set here.

The clumps known to exist within the Bar do possess higher H densities (i.e., 1010 cm) and column densities (i.e., 10 cm; Lis & Schilke, 2003) and would provide the necessary FUV shielding to allow O to reach its full predicted abundance. Such conditions help to reconcile observation and theory in two ways. First, as shown in Fig. 6, the predicted total O column densities decrease with higher H densities. Thus, the total O column density is predicted to be lower if the O emission arises primarily from within the dense clumps rather than the surrounding lower density interclump medium. Second, interferometric observations indicate that the dense clumps within the Bar typically subtend angles of between 4 and 8 (see, for example, Lis & Schilke, 2003), and thus provide a natural explanation for why the beam filling factor of O emission could be less than unity. However, whether the correct explanation for what we observe is that O emission originates preferentially within the dense clumps, and is suppressed within the A7 interclump medium, and with both gas components governed by the processes described in Hollenbach etal. (2009), will depend on how well this model reproduces the wealth of new lines being detected toward the Orion Bar by Herschel.

6 Summary

1. We have conducted a search for O toward the Orion Bar, carrying out deep integrations around the frequencies of the 31 and 53 transitions at 487 GHz and 774 GHz, respectively. Neither line was detected, but sufficiently sensitive limits on their integrated intensities were obtained to test current models of molecular gas exposed to high fluxes of FUV radiation – i.e., 10. In particular, we infer a total face-on O column density of 410 cm, assuming a Bar geometry in which the line-of-sight depth is more than 4 times greater than its face-on dimension. This column density is at least 2 times less than that predicted by the model of Hollenbach etal. (2009) for the densities, temperatures, and  appropriate to the Orion Bar.

2. The discrepancy between the model predictions and our observations would be reduced, if not eliminated, if the adsorption energy of atomic oxygen to wate-ice were greater than 800K, and possibly as high as 1600K. A lower value for the photodesorption yield for HO would help, but is not supported by fits to other astronomical data or recent theoretical calculations and laboratory measurements. A lower grain cross-sectional area per H, such as might occur through grain coagulation, radiation pressure, or the preferential destruction of small grains, would lower the O column density, but it is unclear whether these grain properties apply within the Orion Bar.

3. If the total face-on depth of the interclump medium within the Orion Bar corresponds to an A7, then photodissociation will reduce the O column density to values below our detection limit. Clumps embedded within the Bar would offer sufficient shielding to enable the buildup of higher O abundances and column densities in accord with model predictions, while the small filling factor of these clumps would reduce the O line flux to levels consistent with our upper limits.

4. If the total face-on depth of the interclump medium within the Orion Bar corresponds to an A8, it remains possible that most of the O emission may have been missed. In particular, since the gas density affects the angular separation between the ionization front and the face-on depth into the Bar at which the O abundance is predicted to peak, interclump H densities much different than the assumed value of 10 cm could result in the position of peak O abundance and column density occurring to either the northwest or southeast of the position we selected.

Only further modeling, including predictions for other species, can establish which, if any, of the above possibilities is most likely to resolve the present puzzle.

HIFI has been designed and built by a consortium of institutes and university departments from across Europe, Canada and the United States under the leadership of SRON Netherlands Institute for Space Research, Groningen, The Netherlands, and with major contributions from Germany, France and the US. Consortium members are: Canada: CSA, U. Waterloo; France: CESR, LAB, LERMA, IRAM; Germany: KOSMA, MPIfR, MPS; Ireland, NUI Maynooth; Italy: ASI, IFSI-INAF, Osservatorio Astrofisico di Arcetri-INAF; Netherlands: SRON, TUD; Poland: CAMK, CBK; Spain: Observatorio Astronómico Nacional (IGN), Centro de Astrobiologá (CSIC-INTA). Sweden: Chalmers University of Technology - MC2, RSS & GARD; Onsala Space Observatory; Swedish National Space Board, Stockholm University - Stockholm Observatory; Switzerland: ETH Zurich, FHNW; USA: Caltech, JPL, NHSC. We also acknowledge the effort that went into making critical spectroscopic data available through the Jet Propulsion Laboratory Molecular Spectroscopy Data Base (, the Cologne Database for Molecular Spectroscopy ( and Müller etal. 2005) and the Leiden Atomic and Molecular Database ( and Schöier etal. 2005). Finally, it is a pleasure to acknowledge useful discussions with Dr. Edwin Bergin.

Support for this work was provided by NASA through an award issued by JPL/Caltech.

TABLE 1. Summary of Observations

*[1.5mm] Gaussian Fit Parameters Rest  Observing Integration T LSR Line Integrated Species Transition Frequency Mode  Time Amplitude Center FWHM Intensity (GHz) (hrs) (K) (km s)  (km s)  (K-km s) HCl 10 F3/23/2 485.413 sc 1.16   0.055 10.56 2.47 0.15 F5/23/2 485.418 sc 1.16   0.076 10.56 2.47 0.20 F1/23/2 485.421 sc 1.16   0.030 10.57 2.47 0.08  SO 21/219/2    486.837 sc 1.85   0.029 10.77 2.28 0.07 1/2, e  SO 21/219/2 487.212 sc 1.85   0.027 10.99 1.86 0.05 1/2, f  O 31 487.249 sc 1.85 0.008  CS 109 489.751 sc 0.46 0.46 10.58 1.78 0.87  CO 76 771.184 sp 1.15 27.04 10.67 2.24 64.48  O 53 773.840 sc, sp 10.91 0.007  CH N98 785.802 sc, sp 10.91 0.34 10.76 2.35 0.84 19/217/2 F98  CH N98 785.865 sc, sp 10.91 0.30 10.77 2.35 0.75 17/215/2 F98  CO 76 786.281 sc, sp 10.91 1.19 10.62 1.76 2.23

0         NRAO-recommended rest frequency.   sc: spectral scan observation;  sp: single point observation.
0       3 upper limit.

Fig. 1: Position of the HIFI 44.7 and 28.2 beams at 487 GHz and 774 GHz, respectively, superposed on an HST image of the Orion nebula [O’Dell & Wong, 1996]. Also shown are contours of CO 32 integrated intensity for a portion of a larger map obtained by Lis & Schilke [2003], with intensities in Kkm s noted. The HIFI beams are centered at 0535206, 052514 (J2000), toward the surface layers of the FUV-illuminated Orion Bar where the O emission is predicted to peak.

Fig. 2: Averaged H and V polarization spectra obtained in HIFI Band 1a toward the Orion Bar, ordered by rest frequency, with the Gaussian fits given in Table 1 superposed in red. Also indicated is the energy of the upper level for each transition, in Kelvins. The HCl line is a blend of three, partially resolved, hyperfine components (see Table 1). An LSR velocity of 10.7 km s is denoted with a vertical dashed line.

Fig. 3: Averaged H and V polarization spectra obtained in HIFI Band 2b toward the Orion Bar, ordered by rest frequency, with the Gaussian fits given in Table 1 superposed in red. Also indicated is the energy of the upper level for each transition, in Kelvins. An LSR velocity of 10.7 km s is denoted with a vertical dashed line. The frequency of the CO 76 transition was near the edge of the band and, thus, the spectrum does not cover the full LSR velocity range of the other lines. The O spectrum has been truncated to remove features in the other sideband.

Fig. 4: Contours of integrated antenna temperature, in K km s, under a Gaussian line profile versus line-of-sight O column density and gas temperature assuming an aperture efficiency of 0.7. The shaded area bounds the range of upper limits to the O 487 GHz integrated intensity assuming the intrinsic O line FWHM is 1.8 km s (0.0150 Kkm s) or 2.5 km s (0.0209 Kkm s).

Fig. 5: Contours of integrated antenna temperature, in K km s, under a Gaussian line profile versus line-of-sight O column density and gas temperature assuming an aperture efficiency of 0.7. The shaded area bounds the range of upper limits to the O 774 GHz integrated intensity assuming the intrinsic O line FWHM is 1.8 km s (0.0126 Kkm s) or 2.5 km s (0.0175 Kkm s).

Fig. 6: Predicted total O column density perpendicular to the ionization front as a function of  and (H2H) for HO photodesorption yields of 10 (solid lines) and 10 (dashed line). The results shown assume a cloud thickness sufficient to encompass the zone of peak abundance [after Hollenbach etal., 2009]. The range of  that applies to the Orion Bar is shown in the shaded region. The horizontal dotted line denotes the upper limit to the O column density established here, i.e., 1.510 cm, divided by a geometrical enhancement factor of 4.

Fig. 7: Top panel: Abundance of O as a function of face-on depth into a cloud, measured in A, for a cloud with (H2H)10 cm and 10 cm exposed to a FUV field of 10. This result was computed using the model described in Hollenbach etal. [2009] assuming their “standard” model parameters, except for those noted here. An HO photodesorption yield of 10 is assumed. The gas and dust temperatures throughout the cloud are calculated self-consistently in the Hollenbach et al. code, which predicts a gas temperature of 33K, and a dust temperature of 42K, at the depth of the peak O abundance above. Bottom panel: Cumulative face-on column density of O integrated from the cloud surface to a given depth, in A, for the abundance profile shown in the top panel.


  1. Allers, K.N., Jaffe, D.T, Lacy, J.H., Draine, B.T., Richter, M.J. 2005, Ap. J., 630, 368
  2. Andersson, S., & van Dishoeck, E.F. 2008, A&A, 491, 907
  3. Arasa, C., Andersson, S., Cuppen, H.M., van Dishoeck, E.F., & Kroes, G.J. 2011, J. Chem. Phys., 134, 164503
  4. Batrla, W., & Wilson, T. L. 2003, A&A, 408, 231
  5. Cesarsky, D., Jones, A. P., Lequeux, J., & Verstraete, L. 2000, A&A, 358, 708
  6. de Graauw, T., etal. 2010, A&A, 518, L6
  7. Drouin, B. J., Yu, S., Miller, C. E. etal. 2010, JQSRT, 111, 1167
  8. Goicoechea, J.R., Joblin, C., Contursi, A., Berné, O., Cernicharo, J., Gerin, M., Le Bourlot, J., Bergin, E.A., Bell, T.A., Röllig, M. 2011, A&A, 530, L16
  9. Goldsmith, P. F., Bergin, E. A., & Lis, D. C. 1997, Ap. J., 491, 615
  10. Goldsmith, P. F., & Langer, W.D. 1978, Ap. J., 222, 881
  11. Goldsmith, P. F., Liseau, R., Bell, T. A., etal. 2011, Ap. J., 737, 96
  12. Gorti, U., & Hollenbach, D.J. 2002, Ap. J., 573, 215
  13. Habing, H. J. 1968, Bull. Astron. Inst. Netherlands, 19, 421
  14. Hasegawa, T. I., & Herbst, E. 1993, M.N.R.A.S., 261, 83
  15. Herrmann, F., Madden, S. D., Nikola, T., Poglitsch, A., Timmermann, R., Geis, N., Townes, C. H., & Stacey, G. J. 1997, Ap. J., 481, 343
  16. Hirota, T., etal. 2007, Publ. Astron. Soc. Japan, 59, 897
  17. Hogerheijde, M. R., Jansen, D. J., & Van Dishoeck, E. F. 1995, A&A, 294, 792
  18. Hollenbach, D.J., Kaufman, M. J., Bergin, E.A., & Melnick, G.J. 2009, Ap. J., 690, 1497
  19. Jansen, D.J., Spaans, M., Hogerheijde, M.R., Van Dishoeck, E.F. 1995, A&A, 303, 541
  20. Kim, M. K., etal. 2008, Publ. Astron. Soc. Japan, 60, 991
  21. Lique, F. 2010, J. Chem. Phys., 132, 044311
  22. Lis, D. C., & Schilke, P. 2003, Ap. J., 597, L145
  23. Lis, D. C., Pearson, J. C., Neufeld, D. A., etal. 2010, A&A, 521, L9
  24. Liseau, R., Goldsmith, P. F., Larsson, B., etal. 2012, submitted
  25. Mathis, J. S., Rumpl, W., & Nordsieck, K. H. 1977, Ap. J., 217, 425
  26. Marconi, A., Testi, L., Natta, A., & Walmsley, C. M. 1998, A&A, 330, 696
  27. Menten, K. M., Reid, M. J., Forbrich, J., & Brunthaler, A. 2007, A&A, 474, 515
  28. Müller, H. S. P., Schöder, F., Stutzki, J., & Winnewisser, G. 2005, J. Mol. Struct., 742, 215
  29. Neufeld, D. A., Lepp, S., & Melnick, G. J. 1995, Ap. J. Suppl., 100, 132
  30. Neufeld, D. A., Schilke, P., Menten, K. M., Wolfire, M. G., Black, J. H., Schuller, F., Müller, H. S. P., Thorwirth, S., Güsten, R., & Philipp, S. 2006, A&A, 454, L37
  31. Neufeld, D. A., Roueff, E., Snell, R. L. etal. 2011, submitted to Ap. J. 
  32. Öberg, K. I., Linnartz, H., Visser, R., & van Dishoeck, E. F. 2009, Ap. J., 693, 1209
  33. O’Dell, C. R., & Wong, S. K. 1996, A. J., 111, 846
  34. Ott, S. 2010, ASP Conf. Ser. 434, Astronomical Data Analysis Software and Systems XIX, ed. Y. Mizuno, K. I. Morita, & M. Ohishi (San Francisco, CA: ASP), 139
  35. Pellegrini, E. W., Baldwin, J. A., Ferland, G. J., Shaw, G., & Heathcote, S. 2009, Ap. J., 693, 285
  36. Pickett, H. M., Poynter, R. L., Cohen, E. A., Delitsky, M. L., Pearson, J. C., & Müller, H. S. P. 1998, ”Submillimeter, Millimeter, and Microwave Spectral Line Catalog,” J. Quant. Spectrosc. & Rad. Transfer, 60, 883
  37. Roelfsema, P. R., Helmich, F. P., Teyssier, D., etal. 2012, A&A, 537, A17
  38. Schöier, F. L., van der Tak, F. F. S., van Dishoeck, E. F., & Black, J. H. 2005, A&A, 432, 369
  39. Shaw, G., Ferland, G. J., Henney, W. J., Stancil, P. C., Abel, N. P., Pellegrini, E. W., Baldwin, J. A., & van Hoof, P. A. M. 2009, Ap. J., 701, 677
  40. Sternberg, A. & Dalgarno, A. 1995, Ap. J. Suppl., 99, 565
  41. Simon, R., Stutzki, J., Sternberg, A., & Winnewisser, G. 1997, A&A, 327, L9
  42. Walmsley, C. M., Natta, A., Oliva, E., & Testi, L. 2000, A&A, 364, 301
  43. Westley, M. S., Baragiola, R. A., Johnson, R. E., & Baratta, G. A. 1995a, Nature, 373, 405
  44. Westley, M. S., Baragiola, R. A., Johnson, R. E., & Baratta, G. A. 1995b, Planet. Space Sci., 43, 1311
  45. Wyrowski, F., Schilke, P., Hofner, P., & Walmsley, C. M. 1997, Ap. J., 487, L171
  46. Young Owl, R. C., Meixner, M. M., Wolfire, M., Tielens, A. G. G. M., & Tauber, J. 2000, Ap. J., 540, 886
Comments 0
Request comment
You are adding your first comment
How to quickly get a good reply:
  • Give credit where it’s Due by listing out the positive aspects of a paper before getting into which changes should be made.
  • Be specific in your critique, and provide supporting evidence with appropriate references to substantiate general statements.
  • Your comment should ultimately help the author improve the paper.
  • Remember, the better we can be at sharing our knowledge with each other, the faster we can move forward.
The feedback must be of minumum 40 characters
Add comment
Loading ...
This is a comment super asjknd jkasnjk adsnkj
The feedback must be of minumum 40 characters
The feedback must be of minumum 40 characters

You are asking your first question!
How to quickly get a good answer:
  • Keep your question short and to the point
  • Check for grammar or spelling errors.
  • Phrase it like a question
Test description